Organotypic cultures are tissue culture models that mimic in vivo tissue architecture through manipulation of epithelial and stromal cells within and on top of an extracellular matrix. These models incorporate aspects of cell-matrix and epithelial-stromal interactions that cannot be evaluated in monolayer cultures. They can be used to evaluate monoclonal colonies, whereas individual xenograft tumors in animals are established from millions of cells. Colonies in organotypic culture therefore, more accurately mimic tumorigenesis than animal xenografts. Organotypic cultures can be more readily genetically and chemically manipulated than animal models, allowing increased numbers of experiments and less cost. They cannot replace animal models however, because they do not incorporate metabolic, physiologic and immunologic effects. Thus organotypic cultures can potentially be used to screen and improve chemoprevention agents prior to testing in animal models. In addition, they provide opportunities to evaluated hypothesis driven research on carcinogenesis and chemoprevention. We developed organotypic models of normal human endometrium, benign human ovarian tissue, and human borderline ovarian tumors of low malignant potential (LMP). The objective of this project is to further develop these organotypic models into models of carcinogenesis that can be used to study the processes of carcinogenesis and chemoprevention. The experiments proposed will evaluate the interaction of aromatic hydrocarbons, estrogen, progesterone, tamoxifen, differentiation and adult stem cells in endometrial and ovarian carcinogenesis. Our future goals will be to validate these models and to use them in our efforts to develop low-toxicity retinoids, heteroarotinoids, for chemoprevention. This research will enhance our understanding of the process of carcinogenesis and chemoprevention, which could ultimately translate into strategies and pharmaceuticals for prevention of cancer.